MADARS: A Method of Multi-Attributes Generalized Randomization Privacy Preserving
Guo Xiao-li, Jiajia Zhang, Zhao Yang Qu, Yongwen Wang, Ping Yi Guo · International Journal of Multimedia and Ubiquitous Engineering · 2015
With the extending in the domain of data mining application, the research of the privacy preserving in data mining based on k-anonymity becomes more and more concerned.The key point of the research is to protect data while it's still effective in data mining.However, these methods result in too much information loss, and Multi-dimension bucketization does not generalize quasi-identifiers, which make the anonymized data easy to suffer from linking attack.To overcome these drawbacks, we put forward the thought of data classification processing, named MADARS.MADARS first deal with identifier attributes based on MA-Datafly and then randomized multi-sensitive attributes.Experiment results show that compared with the widely used algorithm Incognito the method can greatly improve the efficiency of the privacy protection while reduce the less personal information and the effectiveness of data is greatly increased.This work is supported by National Natural Science Foundation of China (No.51277023).